CCLHDME R Workshop

Welcome to the CCLHDME R Workshop!

• Housekeeping
  • Ensure you have both R and R Studio installed!
  • Ensure you are connected to the WIFI network

• Introductions

• Overview of workshop
  • What we will do…
  • What we won’t do…
  • Flow of the workshop…

Agenda

• Orientation to R Studio and R Programming
• Basic Data Management, Tidying and Transformation
• Mapping
• Plotting
• Census API and Mapping
• Shiny

Workshop Setup

• 1. Download zip file that has been sent out/posted to pcloud
• 2. Open R Studio
• 3. Navigate to File and Click
• 4. Under File, Select Open File
• 5. When the file browser opens, Select CCLHDME_R_workshop_script
• 6. Click Open

R Studio IDE

• Now you will see the R Studio IDE with the script open
• There are multiple windows in the IDE, all with important uses
• 

Object Oriented Programming

• Everything in R is an object, including character and numeric vectors, data “frames”, lists, functions, charts and graphs, and everything else
• This is one of the beauties of R, allowing for simple and complex integration of data, analysis, and outputs
• Each object has a unique name and has a class
• In R, you assign objects with names with the operator " <- " (you can also use " = “, but it is generally better to use” <- " )

my_number <- 8
my_number

## [1] 8

my_vector <- c(1,1,5,10,50)
my_vector

## [1]  1  1  5 10 50

my_character <- "awesome"
my_character

## [1] "awesome"

my_boolean <- TRUE
my_boolean

## [1] TRUE

my_dataframe <- cbind(my_vector,my_vector^2)
my_dataframe

##      my_vector     
## [1,]         1    1
## [2,]         1    1
## [3,]         5   25
## [4,]        10  100
## [5,]        50 2500

Classes

• R differentiates between different types of data
• you will likely only ever deal with five flavors of data in R, which are stored as three data types

class(my_number)

## [1] "numeric"

class(pi)

## [1] "numeric"

class(my_boolean)

## [1] "logical"

class(FALSE)

## [1] "logical"

class("This object is made of characters")

## [1] "character"

class(as.Date("2015-07-27"))  

## [1] "Date"

• The “factor” class will be discussed later–there are other classes too

A note on dates in R

• R stores dates as the number of days since the epoch (1 Jan 1970)
• Regular arithmetic operations can be conducted on dates
• The lubridate package (part of tidyverse) has many date-related functions that make working with dates much easier
• The function as.Date() converts character representations of a date to an object of class “Date” representing calendar dates

class(as.Date("2015-07-27"))  

## [1] "Date"

Functions

• A function is a set of statements that perform a specific task (they can be much like “procs” in SAS)
• In R, there are many functions built in to base R and packages
• Functions are called by name
• Data are explicitly passed to functions
• Functions all have the same basic structure: function(arguments) where the arguments are other objects (i.e. data)

• When you call a function, it looks at the classes of the things you are calling it on to figure out how to behave

• there are many simple and obvious functions like mean(), median(), min()
  

• c() function is used to create a vector, which is an ordered group of the same kind of data

• Functions are used to do a lot of different things!

mean(my_vector)

## [1] 13.4

max(my_vector)

## [1] 50

paste0(my_character,"...", my_character)

## [1] "awesome...awesome"

sum(1,2,3,3,4,5)

## [1] 18

mean(c(1,2,3,4,5))

## [1] 3

making your own functions is a more advanced topic, but ultimately is essential for using R beyound the basics

myHypotenuse <- function(side1,side2) {sqrt(side1^2 + side2^2)}

myHypotenuse(10,20)

## [1] 22.36068

myHypotenuse(20,35)

## [1] 40.31129

Operators

• Operators are mathematical and logical tools that are used often for analysis and data processing
• Mathematical:
• ‘+’, ‘-’, ’*‘,’/‘,’^’
• Logical:
• ‘==’ (equal),‘>’, ‘<’, ‘<=’, ‘>=’, ‘%in%’
• and, any of the logical operators can be preceded with an “!” to be “NOT”…

• more information on operators in R HERE https://www.datamentor.io/r-programming/operator/

• In some cases, some operators will only work with specific classes (e.g. “bob”*“bill” does not make sense, and will generate an error)

20/5

## [1] 4

my_number*my_number

## [1] 64

2 > 3

## [1] FALSE

5^2 == 25

## [1] TRUE

"joe" %in% c("bill","sue","joe")

## [1] TRUE

Directories

• In R you are always in a directory
• Your actions are all relative to that directory
• When you work in “projects”, the directory is always set to where your project is stored

getwd()

## [1] "C:/Users/rfisher/Desktop/CCLHDME_R_workshop"

• BUT, you can also set your directory to some other location using setwd()

#setwd("I:/CCLHDME/R proposal/CCLHDME_R_workshop")

Testing data types

• you can test data types (i.e. ask logical equivalency questions)

is.character(my_number)

## [1] FALSE

is.numeric(my_number)

## [1] TRUE

Changing data types

• you can change data types (if it makes sense!!!)

my_number_as_character <- as.character(my_number)
my_character_as_number <- as.numeric(my_character)
print(my_number_as_character)

## [1] "8"

print(my_character_as_number)

## [1] NA

-trying to coerce types can lead to weird behavior

Data Structures

• there are five kinds of data structures in R, but you will probably only ever use three of these
  • vector
  • list
  • dataframe (in Tidyverse=Tibble)
• a vector is an ordered group of the same kind of data
• a list is an ordered group of things that are not of the same type
• a dataframe is just a list of equal-length vectors
  • What we are all probably used to!
  • Tabular data, i.e. columns and rows

Packages in R

• Availability of all sorts of “packages” beyond base R is one of the things the makes R so powerful and useful. There are many specialty packages for epidemiology, for surveillance, for data presentation, and for almost anything you might want to do with data
• Open source software leads to a lot of package development
• Download from repositories (online and openly accessible)
  • CRAN: Official repository—packages here need to pass several tests that ensure the package is following CRAN policies.
  • Github: not R specific, but is very popular repository for open source projects. No review process associated with this repository.

Get Documentation on R packages

• Learn to read package documentation and find tutorials/use cases
  • See resource list (not exhaustive so dig around the internet!)
  • Can use a search engine to find documentation and examples
  • https://cran.r-project.org/web/packages/available_packages_by_name.html
  • https://stackoverflow.com/questions/tagged/r
• From script/console:

help(package = "tidyverse")

Installing packages

• Install packages using the manual installer
• Install in script/console using: install.packages(“package_name”)

#install.packages("tidyverse") 

Loading or Calling packages in R session

• Any package that is needed in an R session must be loaded in that session
  
• Load into environment using: library(package_name)

library(tidyverse)

Tidyverse

• Tidyverse is a set of packages that makes many aspects of data manipulation and analysis (“data wrangling” in tidy-speak) much easier
• Tidyverse includes many packages that you’re likely to use in every day data analyses
• The tidyverse collection of packages “share an underlying design philosophy, grammar and data structures” of tidy data
  • Learn more at: https://www.tidyverse.org/packages/
  • And: https://www.tidyverse.org/
  • And: https://tidyverse.tidyverse.org/
  • And: https://r4ds.had.co.nz/

What is Tidy data?

• 
• Check out Rstudio Cheatsheets (there’s one on Tidyr, other tidyverse packages and more): https://www.rstudio.com/resources/cheatsheets/

On to working with data!

• We will be working with data form the CHHS open data portal
• In your browser, navigate to: https://data.chhs.ca.gov/dataset/infectious-disease-cases-by-county-year-and-sex
• 

Download CSV File

• Select Download
• Save file in our data sub-directory
• 

Import data from the CSV

• Can import data manually through the “import dataset” utility under File Menu
• We will import the CSV file using read_csv()
• Remember we want to name an object that we will store the data to using <-

ca_id_data <- read_csv("./data/infectious-disease-cases-by-county-year-and-sex-2-27-19.csv")

Importing from other file types

• Read semi-colon delimited files using read_csv2()
• Read files with any delimiter using read_delim()
• Read tab-delimited files using read_tsv()
• Read fixed width files using read_fwf()
• Read SPSS, Stata and SAS files with Haven (part of “tidyverse”) or foreign package
  • Need to install and call
• Read excel files with Readxl package
  • Need to install and call

Import data directly from the Open Data Portal using the API (Application Program Interface)

• We can also pull the data directly from CHHS through the API using the url

Now we pull the data from the API

ca_id_data_api <- read_csv("https://data.chhs.ca.gov/dataset/391d75bf-00fc-4fd7-b00d-9bd16bbe01c0/resource/741f25e1-db50-436a-a6a9-7b840176edbf/download/infectious-disease-cases-by-county-year-and-sex-2-27-19.csv")
str(ca_id_data_api)

## Classes 'spec_tbl_df', 'tbl_df', 'tbl' and 'data.frame': 132927 obs. of  10 variables:
##  $ Disease   : chr  "Amebiasis" "Amebiasis" "Amebiasis" "Amebiasis" ...
##  $ County    : chr  "Solano" "Marin" "Kern" "Tulare" ...
##  $ Year      : num  2005 2005 2010 2001 2014 ...
##  $ Sex       : chr  "Total" "Male" "Female" "Female" ...
##  $ Count     : num  0 4 1 1 17 2 0 0 3 0 ...
##  $ Population: num  410570 121710 404863 186935 5127242 ...
##  $ Rate      : num  0 3.287 0.247 0.535 0.332 ...
##  $ CI. lower : num  0 0.895 0.006 0.014 0.193 ...
##  $ CI.upper  : num  0.898 8.415 1.376 2.98 0.531 ...
##  $ Unstable  : chr  "-" "*" "*" "*" ...
##  - attr(*, "spec")=
##   .. cols(
##   ..   Disease = col_character(),
##   ..   County = col_character(),
##   ..   Year = col_double(),
##   ..   Sex = col_character(),
##   ..   Count = col_double(),
##   ..   Population = col_double(),
##   ..   Rate = col_double(),
##   ..   `CI. lower` = col_double(),
##   ..   CI.upper = col_double(),
##   ..   Unstable = col_character()
##   .. )

A brief note on “Factors” in R (advanced topic)

• Factors are variables that generally “look like” character string variables, but are actually stored internally as integers, with the character strings being just the ‘labels’
• Factors can be very useful, but for many basic “epi/data management” purposes (except multivariate regression), they can be easier to use if they are converted into “regular” character variables (e.g. my_string_variable <- as.character(my_factor_variable))
  
• when reading in .csv data using “read.csv()”, all character sting variables are read in by default as factors. To avoid this, you can use the “as.is=TRUE” argument to the read.csv function, or you can use read_csv (from the tidyverse readr() package)
• factor() function creates factors, which can be useful in a number of situations

ca_id_data_api_FACTOR <- read.csv("https://data.chhs.ca.gov/dataset/391d75bf-00fc-4fd7-b00d-9bd16bbe01c0/resource/741f25e1-db50-436a-a6a9-7b840176edbf/download/infectious-disease-cases-by-county-year-and-sex-2-27-19.csv")
str(ca_id_data_api)

## Classes 'spec_tbl_df', 'tbl_df', 'tbl' and 'data.frame': 132927 obs. of  10 variables:
##  $ Disease   : chr  "Amebiasis" "Amebiasis" "Amebiasis" "Amebiasis" ...
##  $ County    : chr  "Solano" "Marin" "Kern" "Tulare" ...
##  $ Year      : num  2005 2005 2010 2001 2014 ...
##  $ Sex       : chr  "Total" "Male" "Female" "Female" ...
##  $ Count     : num  0 4 1 1 17 2 0 0 3 0 ...
##  $ Population: num  410570 121710 404863 186935 5127242 ...
##  $ Rate      : num  0 3.287 0.247 0.535 0.332 ...
##  $ CI. lower : num  0 0.895 0.006 0.014 0.193 ...
##  $ CI.upper  : num  0.898 8.415 1.376 2.98 0.531 ...
##  $ Unstable  : chr  "-" "*" "*" "*" ...
##  - attr(*, "spec")=
##   .. cols(
##   ..   Disease = col_character(),
##   ..   County = col_character(),
##   ..   Year = col_double(),
##   ..   Sex = col_character(),
##   ..   Count = col_double(),
##   ..   Population = col_double(),
##   ..   Rate = col_double(),
##   ..   `CI. lower` = col_double(),
##   ..   CI.upper = col_double(),
##   ..   Unstable = col_character()
##   .. )

• factor() function creates factors

class(factor(c('Salmonellosis', 'Campylobacteriosis', 'Amebiasis'))) 

## [1] "factor"

Exploring our data

• Now that we have our data loaded as a dataframe in R, we want to explore its features and characteristics
• We can view the dataframe by manualy clicking the data object in the global environment
• We can also use a function to open the dataframe: View()

View(ca_id_data_api)

Explore data using str()

str(ca_id_data_api)

## Classes 'spec_tbl_df', 'tbl_df', 'tbl' and 'data.frame': 132927 obs. of  10 variables:
##  $ Disease   : chr  "Amebiasis" "Amebiasis" "Amebiasis" "Amebiasis" ...
##  $ County    : chr  "Solano" "Marin" "Kern" "Tulare" ...
##  $ Year      : num  2005 2005 2010 2001 2014 ...
##  $ Sex       : chr  "Total" "Male" "Female" "Female" ...
##  $ Count     : num  0 4 1 1 17 2 0 0 3 0 ...
##  $ Population: num  410570 121710 404863 186935 5127242 ...
##  $ Rate      : num  0 3.287 0.247 0.535 0.332 ...
##  $ CI. lower : num  0 0.895 0.006 0.014 0.193 ...
##  $ CI.upper  : num  0.898 8.415 1.376 2.98 0.531 ...
##  $ Unstable  : chr  "-" "*" "*" "*" ...
##  - attr(*, "spec")=
##   .. cols(
##   ..   Disease = col_character(),
##   ..   County = col_character(),
##   ..   Year = col_double(),
##   ..   Sex = col_character(),
##   ..   Count = col_double(),
##   ..   Population = col_double(),
##   ..   Rate = col_double(),
##   ..   `CI. lower` = col_double(),
##   ..   CI.upper = col_double(),
##   ..   Unstable = col_character()
##   .. )

Explore data using print()

print(ca_id_data)

## # A tibble: 132,927 x 10
##    Disease County  Year Sex   Count Population  Rate `CI. lower` CI.upper
##    <chr>   <chr>  <dbl> <chr> <dbl>      <dbl> <dbl>       <dbl>    <dbl>
##  1 Amebia~ Solano  2005 Total     0     410570 0           0        0.898
##  2 Amebia~ Marin   2005 Male      4     121710 3.29        0.895    8.41 
##  3 Amebia~ Kern    2010 Fema~     1     404863 0.247       0.006    1.38 
##  4 Amebia~ Tulare  2001 Fema~     1     186935 0.535       0.014    2.98 
##  5 Amebia~ Los A~  2014 Fema~    17    5127242 0.332       0.193    0.531
##  6 Amebia~ Napa    2006 Total     2     131920 1.52        0.184    5.48 
##  7 Amebia~ Marip~  2011 Total     0      18237 0           0       20.2  
##  8 Amebia~ Shasta  2006 Male      0      85947 0           0        4.29 
##  9 Amebia~ Sutter  2003 Male      3      41653 7.20        1.48    21.0  
## 10 Amebia~ Lake    2001 Male      0      29796 0           0       12.4  
## # ... with 132,917 more rows, and 1 more variable: Unstable <chr>

Explore data using head()

head(ca_id_data, 10)

## # A tibble: 10 x 10
##    Disease County  Year Sex   Count Population  Rate `CI. lower` CI.upper
##    <chr>   <chr>  <dbl> <chr> <dbl>      <dbl> <dbl>       <dbl>    <dbl>
##  1 Amebia~ Solano  2005 Total     0     410570 0           0        0.898
##  2 Amebia~ Marin   2005 Male      4     121710 3.29        0.895    8.41 
##  3 Amebia~ Kern    2010 Fema~     1     404863 0.247       0.006    1.38 
##  4 Amebia~ Tulare  2001 Fema~     1     186935 0.535       0.014    2.98 
##  5 Amebia~ Los A~  2014 Fema~    17    5127242 0.332       0.193    0.531
##  6 Amebia~ Napa    2006 Total     2     131920 1.52        0.184    5.48 
##  7 Amebia~ Marip~  2011 Total     0      18237 0           0       20.2  
##  8 Amebia~ Shasta  2006 Male      0      85947 0           0        4.29 
##  9 Amebia~ Sutter  2003 Male      3      41653 7.20        1.48    21.0  
## 10 Amebia~ Lake    2001 Male      0      29796 0           0       12.4  
## # ... with 1 more variable: Unstable <chr>

Explore data using tail()

tail(ca_id_data, 10)

## # A tibble: 10 x 10
##    Disease County  Year Sex   Count Population  Rate `CI. lower` CI.upper
##    <chr>   <chr>  <dbl> <chr> <dbl>      <dbl> <dbl>       <dbl>    <dbl>
##  1 Yersin~ Amador  2003 Fema~     0      16572     0           0   22.3  
##  2 Yersin~ Colusa  2012 Male      0      11200     0           0   32.9  
##  3 Yersin~ Kings   2008 Total     0     151834     0           0    2.43 
##  4 Yersin~ Marip~  2004 Fema~     0       8791     0           0   42.0  
##  5 Yersin~ Lake    2004 Fema~     0      31503     0           0   11.7  
##  6 Yersin~ Monte~  2012 Fema~     0     205765     0           0    1.79 
##  7 Yersin~ Imper~  2013 Male      0      92756     0           0    3.98 
##  8 Yersin~ Nevada  2013 Total     0      97727     0           0    3.78 
##  9 Yersin~ Humbo~  2011 Total     0     135332     0           0    2.73 
## 10 Yersin~ Sacra~  2007 Total     0    1388086     0           0    0.266
## # ... with 1 more variable: Unstable <chr>

Explore data using summary()

summary(ca_id_data)

##    Disease             County               Year          Sex           
##  Length:132927      Length:132927      Min.   :2001   Length:132927     
##  Class :character   Class :character   1st Qu.:2004   Class :character  
##  Mode  :character   Mode  :character   Median :2008   Mode  :character  
##                                        Mean   :2008                     
##                                        3rd Qu.:2012                     
##                                        Max.   :2015                     
##      Count            Population            Rate         
##  Min.   :    0.00   Min.   :     555   Min.   :   0.000  
##  1st Qu.:    0.00   1st Qu.:   28946   1st Qu.:   0.000  
##  Median :    0.00   Median :  124049   Median :   0.000  
##  Mean   :   29.09   Mean   :  833883   Mean   :   3.783  
##  3rd Qu.:    1.00   3rd Qu.:  413167   3rd Qu.:   0.053  
##  Max.   :49975.00   Max.   :39059809   Max.   :2989.961  
##    CI. lower           CI.upper          Unstable        
##  Min.   :   0.000   Min.   :   0.009   Length:132927     
##  1st Qu.:   0.000   1st Qu.:   1.429   Class :character  
##  Median :   0.000   Median :   4.768   Mode  :character  
##  Mean   :   2.793   Mean   :  25.308                     
##  3rd Qu.:   0.002   3rd Qu.:  17.422                     
##  Max.   :2773.224   Max.   :3218.730

Basic data management tasks using tidyr and dplyr

• Now we will cover some basic data management tasks and transformations using Tidyr and Dplyr packages
  • We do not have to call in the package because it is a part of Tidyverse
  • If we wanted to call it in as a standalone package we would use library(tidyr) and library(dplyr)

library(tidyr)
library(dplyr)

Intro to the pipe operator: %>%

• Pipes are a way to chain commands together
• They help make code easier to write and read
• Interpret a pipe as “and then…”

Select Observations

• We may want to select particular observations
• In this case we want to create a dataframe of 2015 data for all population in California

ca_id_data_filtered <- ca_id_data %>% filter(Sex == "Total", Year == 2015, County == "California")
head(ca_id_data_filtered, 15)

## # A tibble: 15 x 10
##    Disease County  Year Sex   Count Population   Rate `CI. lower` CI.upper
##    <chr>   <chr>  <dbl> <chr> <dbl>      <dbl>  <dbl>       <dbl>    <dbl>
##  1 Amebia~ Calif~  2015 Total   309   39059809  0.791       0.705    0.884
##  2 Anapla~ Calif~  2015 Total     4   39059809  0.01        0.003    0.026
##  3 Anthrax Calif~  2015 Total     0   39059809  0           0        0.009
##  4 Babesi~ Calif~  2015 Total     7   39059809  0.018       0.007    0.037
##  5 Botuli~ Calif~  2015 Total     1   39059809  0.003       0        0.014
##  6 Botuli~ Calif~  2015 Total     2   39059809  0.005       0.001    0.018
##  7 Botuli~ Calif~  2015 Total    11   39059809  0.028       0.014    0.05 
##  8 Brucel~ Calif~  2015 Total    30   39059809  0.077       0.052    0.11 
##  9 Campyl~ Calif~  2015 Total  8272   39059809 21.2        20.7     21.6  
## 10 Cholera Calif~  2015 Total     0   39059809  0           0        0.009
## 11 Ciguat~ Calif~  2015 Total     1   39059809  0.003       0        0.014
## 12 Coccid~ Calif~  2015 Total  3060   39059809  7.83        7.56     8.12 
## 13 Creutz~ Calif~  2015 Total    18   39059809  0.046       0.027    0.073
## 14 Crypto~ Calif~  2015 Total   375   39059809  0.96        0.865    1.06 
## 15 Cyclos~ Calif~  2015 Total    15   39059809  0.038       0.021    0.063
## # ... with 1 more variable: Unstable <chr>

Select Variables

• We may want to select particular variables
• In this case we want to create a dataframe of 2015 counts and population only for all population in California

ca_id_data_selected <- ca_id_data_filtered %>% select(Disease, Year, Count, Population)
head(ca_id_data_selected, 15)

## # A tibble: 15 x 4
##    Disease                                            Year Count Population
##    <chr>                                             <dbl> <dbl>      <dbl>
##  1 Amebiasis                                          2015   309   39059809
##  2 Anaplasmosis and Ehrlichiosis                      2015     4   39059809
##  3 Anthrax                                            2015     0   39059809
##  4 Babesiosis                                         2015     7   39059809
##  5 Botulism, Foodborne                                2015     1   39059809
##  6 Botulism, Other                                    2015     2   39059809
##  7 Botulism, Wound                                    2015    11   39059809
##  8 Brucellosis                                        2015    30   39059809
##  9 Campylobacteriosis                                 2015  8272   39059809
## 10 Cholera                                            2015     0   39059809
## 11 Ciguatera Fish Poisoning                           2015     1   39059809
## 12 Coccidioidomycosis                                 2015  3060   39059809
## 13 Creutzfeldt-Jakob Disease and other Transmissibl~  2015    18   39059809
## 14 Cryptosporidiosis                                  2015   375   39059809
## 15 Cyclosporiasis                                     2015    15   39059809

Create new variables

• Let’s create a new variable called incidence_rate

ca_id_data_mutated <- ca_id_data_selected %>% mutate(incidence_rate = round(((Count/Population)*100000), digits = 2))
head(ca_id_data_mutated, 15)

## # A tibble: 15 x 5
##    Disease                             Year Count Population incidence_rate
##    <chr>                              <dbl> <dbl>      <dbl>          <dbl>
##  1 Amebiasis                           2015   309   39059809           0.79
##  2 Anaplasmosis and Ehrlichiosis       2015     4   39059809           0.01
##  3 Anthrax                             2015     0   39059809           0   
##  4 Babesiosis                          2015     7   39059809           0.02
##  5 Botulism, Foodborne                 2015     1   39059809           0   
##  6 Botulism, Other                     2015     2   39059809           0.01
##  7 Botulism, Wound                     2015    11   39059809           0.03
##  8 Brucellosis                         2015    30   39059809           0.08
##  9 Campylobacteriosis                  2015  8272   39059809          21.2 
## 10 Cholera                             2015     0   39059809           0   
## 11 Ciguatera Fish Poisoning            2015     1   39059809           0   
## 12 Coccidioidomycosis                  2015  3060   39059809           7.83
## 13 Creutzfeldt-Jakob Disease and oth~  2015    18   39059809           0.05
## 14 Cryptosporidiosis                   2015   375   39059809           0.96
## 15 Cyclosporiasis                      2015    15   39059809           0.04

Sort

• We may want to sort by one or more variables
• In this case we want to sort our data by descending rate

ca_id_data_mutated <- ca_id_data_mutated %>% arrange(-incidence_rate)
head(ca_id_data_mutated, 15)

## # A tibble: 15 x 5
##    Disease                         Year Count Population incidence_rate
##    <chr>                          <dbl> <dbl>      <dbl>          <dbl>
##  1 Campylobacteriosis              2015  8272   39059809          21.2 
##  2 Salmonellosis                   2015  5568   39059809          14.3 
##  3 HIV                             2015  4947   39059809          12.7 
##  4 Coccidioidomycosis              2015  3060   39059809           7.83
##  5 Shigellosis                     2015  2227   39059809           5.7 
##  6 Giardiasis                      2015  2151   39059809           5.51
##  7 E. coli Other STEC (non-O157)   2015   602   39059809           1.54
##  8 Legionellosis                   2015   452   39059809           1.16
##  9 Cryptosporidiosis               2015   375   39059809           0.96
## 10 E. coli O157                    2015   330   39059809           0.84
## 11 Amebiasis                       2015   309   39059809           0.79
## 12 Vibrio Infection (non-Cholera)  2015   242   39059809           0.62
## 13 Dengue                          2015   135   39059809           0.35
## 14 Listeriosis                     2015   130   39059809           0.33
## 15 Yersiniosis                     2015   122   39059809           0.31

Do it all together! This is the strength of dplyr and the pipe

• We can filter, select, mutate and arrange efficiently by connecting each “step” with the pipe
• Multiple (sort) steps can be on one line, or the can be on separate lines to make the code easier to read (and trouble-shoot)

ca_id_data_cleaned <- ca_id_data %>%
                      filter(Sex == "Total", Year == 2015, County == "California" ) %>%
                      select(Disease,Year,Count,Population) %>%
                      mutate(incidence_rate = round(((Count/Population)*100000), digits = 2)) %>%
                      arrange(-incidence_rate)

Practice data management and using the pipe

• Start with the ca_id_data and create a new dataset called ca_id_data_mydata
• Practice modifying the code from above to make a dataset for your county, for all years, for one disease

Making static tables with Kable

• There are several packages in R to make tables. The Knitr package (part of tidyverse) has a function kable which is good in markdown and for static tables. We will install a package called kableExtra which can be used for stylistic improvement.
• install kable using install.packages(“kableExtra”)

Make a Simple Static Table of top 20 diseases using Kable

• Use the function kable()

cols<-c(1:3,5) #This is creating an object to define which columns from our data we want to use
kable(ca_id_data_cleaned[1:20,cols]) #This is "old school indexing" in R which is largely replaced by subsetting in tidyverse, but is still really important to know

Disease	Year	Count	incidence_rate
Campylobacteriosis	2015	8272	21.18
Salmonellosis	2015	5568	14.26
HIV	2015	4947	12.67
Coccidioidomycosis	2015	3060	7.83
Shigellosis	2015	2227	5.70
Giardiasis	2015	2151	5.51
E. coli Other STEC (non-O157)	2015	602	1.54
Legionellosis	2015	452	1.16
Cryptosporidiosis	2015	375	0.96
E. coli O157	2015	330	0.84
Amebiasis	2015	309	0.79
Vibrio Infection (non-Cholera)	2015	242	0.62
Dengue	2015	135	0.35
Listeriosis	2015	130	0.33
Yersiniosis	2015	122	0.31
Lyme Disease	2015	100	0.26
Malaria	2015	97	0.25
Shiga Toxin Positive Feces (without culture confirmation)	2015	98	0.25
Typhus Fever	2015	78	0.20
Staphylococcus aureus Infection (cases resulting in death or ICU)	2015	59	0.15

A Brief Sidenote on “old school indexing”" in R

• important to understand the basics of indexing
• review “Selecting Vector Elements” and “Matrix subsetting” sections of https://www.rstudio.com/wp-content/uploads/2016/10/r-cheat-sheet-3.pdf

##   x y
## 1 1 a
## 2 2 b
## 3 3 c

## [1] a b c
## Levels: a b c

## [1] 3

Make a Fancy Static Table of top 20 diseases using Kable

kable(ca_id_data_cleaned[1:20,cols],
      align = c('l','c','c','c'),
      col.names = c("Disease","Year","Number of Cases","Incidence Rate"),
      caption = "Top 20 Communicable Diseases  by Incidence Rate in California, 2015") %>%
  kable_styling(bootstrap_options = c("striped", "hover", "condensed", "responsive","bordered"), full_width = T,position = "left", font_size = 12) %>%
  footnote(general = c("Data Retrieved on 5/16/19 from https://data.chhs.ca.gov/dataset/infectious-disease-cases-by-county-year-and-sex"))

Top 20 Communicable Diseases by Incidence Rate in California, 2015

Disease	Year	Number of Cases	Incidence Rate
Campylobacteriosis	2015	8272	21.18
Salmonellosis	2015	5568	14.26
HIV	2015	4947	12.67
Coccidioidomycosis	2015	3060	7.83
Shigellosis	2015	2227	5.70
Giardiasis	2015	2151	5.51
E. coli Other STEC (non-O157)	2015	602	1.54
Legionellosis	2015	452	1.16
Cryptosporidiosis	2015	375	0.96
E. coli O157	2015	330	0.84
Amebiasis	2015	309	0.79
Vibrio Infection (non-Cholera)	2015	242	0.62
Dengue	2015	135	0.35
Listeriosis	2015	130	0.33
Yersiniosis	2015	122	0.31
Lyme Disease	2015	100	0.26
Malaria	2015	97	0.25
Shiga Toxin Positive Feces (without culture confirmation)	2015	98	0.25
Typhus Fever	2015	78	0.20
Staphylococcus aureus Infection (cases resulting in death or ICU)	2015	59	0.15
Note:
Data Retrieved on 5/16/19 from https://data.chhs.ca.gov/dataset/infectious-disease-cases-by-county-year-and-sex

Making interactive tables with DT

• Package DT is great for shiny apps or html reports (exclusively interactive content).
  
• install DT using install.packages(“DT”)

Make simple interactive Table of top 20 diseases using DT

• Use the function datatable()

datatable(ca_id_data_cleaned[1:20,cols])

Make a fancy interactive Table of top 20 diseases using DT

datatable(ca_id_data_cleaned[1:20,cols],
          options = list(columnDefs = list(list(className = 'dt-center', targets = c(1,2,3))),pageLength = 10), 
          class = 'compact stripe nowrap hover order-column row-border', 
          rownames = FALSE,
          colnames = c('Disease' = 1,'Year' = 2,'Number of Cases' = 3,'Incidence Rate' = 4),
          caption = htmltools::tags$caption(style = 'caption-side:top; text-align:left; color:red; font-weight:bold;font-size: 130%;','Top 20 Communicable Diseases  by Incidence Rate in California, 2015'))

Practice making tables

• Create a new dataset called ca_id_data_cleaned_count that is arranged by descending case counts
• Using this new dataset, create a table using either kable (static) or DT (interactive)
• Copy and paste the code above and modify some elements such as the title

Grouping data using group_by()

• Let’s go back to the original dataset
• We want to summarize data across years for all of California for all sexes by Disease
• We have to filter our data first and then group the data by disease

ca_id_data_grouped <- ca_id_data %>%
                      filter(Sex == "Total", County == "California") %>%
                      group_by(Disease) 

Summarize

• Now we can summarize and sort our data by descending rate

ca_id_data_summarized <- ca_id_data_grouped %>%
                         summarise(avg_count      = mean(Count), 
                                   total_count    = sum(Count), 
                                   avg_population = mean(Population), 
                                   avg_rate = mean(Rate)
                                   ) %>%
                         arrange(-avg_rate)       

head(ca_id_data_summarized,15)

## # A tibble: 15 x 5
##    Disease                    avg_count total_count avg_population avg_rate
##    <chr>                          <dbl>       <dbl>          <dbl>    <dbl>
##  1 Hepatitis C, Chronic          29369.      440540      36835584.   80.5  
##  2 Hepatitis B, Chronic          10540.      158104      36835584.   29.1  
##  3 Campylobacteriosis             6093        91395      36835584.   16.5  
##  4 HIV                            5533.       82996      36835584.   15.1  
##  5 Salmonellosis                  4626.       69396      36835584.   12.5  
##  6 Coccidioidomycosis             2896.       43438      36835584.    7.81 
##  7 Giardiasis                     2075.       31127      36835584.    5.66 
##  8 Shigellosis                    1639.       24581      36835584.    4.49 
##  9 Amebiasis                       376         5640      36835584.    1.03 
## 10 Cryptosporidiosis               311.        4662      36835584.    0.841
## 11 E. coli O157                    271.        4068      36835584.    0.735
## 12 E. coli Other STEC (non-O~      203.        2029      37595546.    0.530
## 13 Legionellosis                   175.        2618      36835584.    0.464
## 14 Malaria                         134.        2011      36835584.    0.367
## 15 Vibrio Infection (non-Cho~      135         2025      36835584.    0.362

Practice Grouping and Summarizing

• Create a new dataset called ca_id_data_group_sum_practice that is:
• filtered to Males and Females
• grouped by disease and County
• summarized across all years with average annual counts and average annual rates
• arranged by descending average annual rates

Practice Answer

ca_id_data_group_sum_practice <- ca_id_data %>%
                        filter(Sex == "Total") %>%
                        group_by(County, Disease) %>%
                        summarise(avg_count = mean(Count), 
                                   avg_rate = mean(Rate)
                                   ) %>%
                        arrange(-avg_rate)       

head(ca_id_data_group_sum_practice,15)

## # A tibble: 15 x 4
## # Groups:   County [14]
##    County          Disease              avg_count avg_rate
##    <chr>           <chr>                    <dbl>    <dbl>
##  1 Lassen          Hepatitis C, Chronic     267.      772.
##  2 Del Norte       Hepatitis C, Chronic     179.      637.
##  3 Madera          Hepatitis C, Chronic     429.      303.
##  4 Kings           Hepatitis C, Chronic     437.      297.
##  5 Amador          Hepatitis C, Chronic     106.      283.
##  6 Kern            Hepatitis C, Chronic    1653.      208.
##  7 Tuolumne        Hepatitis C, Chronic     112.      201.
##  8 Imperial        Hepatitis C, Chronic     317.      193.
##  9 Kern            Coccidioidomycosis      1319.      164.
## 10 Humboldt        Hepatitis C, Chronic     209.      158.
## 11 Trinity         Hepatitis C, Chronic      20.8     154.
## 12 Lake            Hepatitis C, Chronic      92.1     146.
## 13 Shasta          Hepatitis C, Chronic     232.      133.
## 14 Mendocino       Hepatitis C, Chronic     116.      133.
## 15 San Luis Obispo Hepatitis C, Chronic     345.      131.

10 MINUTE BREAK

• Next up: We will move on to more advanced applications
• Mapping
• Plots
• Census Data

Now we are going to move on to some more advanced applications

• Mapping
• Plots
• Census Data (maybe)

Mapping

• There’s so much more to dig into and we will go fast through much of this
• There are great tutorials and workshops online that will provide much more detail on geospatial science in R including +coordinate reference system transformation +geocoding +working with raster data

Packages for Mapping

• “sp” package is for classes and methods for spatial data
• “tigris” package is used to download census geographic data
• “tmap” is a great package for making maps and that we will use today – there are many others
• install these packages using install.packages()

#install.packages("sp")
#install.packages("tigris")

library(sp)
library(tigris)

options(tigris_class = "sf")   # options are sp (spatial polygons) or sf (simple features)
options(tigris_use_cache = F)  # set to true to save locally

Access census geographic boundary data

• https://cran.r-project.org/web/packages/tigris/tigris.pdf
• use cb=F for more detailed geographies and cb=T for prettier simpler geography
• State, census tract, “place”, and other shape files are also available with this package

counties2010 <- counties(state = '06', cb = T, year = 2010)

Visualize the boundaires and data quickly using the R plot function

plot(counties2010)

Explore structure of data

• data is in a “simple features” dataframe, and has a special “variable” which has all the shape information

str(counties2010)

## Classes 'sf' and 'data.frame':   58 obs. of  9 variables:
##  $ GEO_ID    : chr  "0500000US06035" "0500000US06049" "0500000US06075" "0500000US06083" ...
##  $ STATE     : chr  "06" "06" "06" "06" ...
##  $ COUNTY    : chr  "035" "049" "075" "083" ...
##  $ NAME      : chr  "Lassen" "Modoc" "San Francisco" "Santa Barbara" ...
##  $ LSAD      : chr  "County" "County" "County" "County" ...
##  $ CENSUSAREA: num  4541.2 3917.8 46.9 2735.1 953.2 ...
##  $ geometry  :sfc_MULTIPOLYGON of length 58; first list element: List of 1
##   ..$ :List of 1
##   .. ..$ : num [1:201, 1:2] -120 -120 -120 -120 -120 ...
##   ..- attr(*, "class")= chr  "XY" "MULTIPOLYGON" "sfg"
##  $ COUNTYFP  : chr  "035" "049" "075" "083" ...
##  $ STATEFP   : chr  "06" "06" "06" "06" ...
##  - attr(*, "sf_column")= chr "geometry"
##  - attr(*, "agr")= Factor w/ 3 levels "constant","aggregate",..: NA NA NA NA NA NA NA NA
##   ..- attr(*, "names")= chr  "GEO_ID" "STATE" "COUNTY" "NAME" ...
##  - attr(*, "tigris")= chr "county"

Explore the data

head(counties2010)

Mapping our Infectious Disease data

• Let’s take another look at our Infectious Disease data

head(ca_id_data)

## # A tibble: 6 x 10
##   Disease County  Year Sex   Count Population  Rate `CI. lower` CI.upper
##   <chr>   <chr>  <dbl> <chr> <dbl>      <dbl> <dbl>       <dbl>    <dbl>
## 1 Amebia~ Solano  2005 Total     0     410570 0           0        0.898
## 2 Amebia~ Marin   2005 Male      4     121710 3.29        0.895    8.41 
## 3 Amebia~ Kern    2010 Fema~     1     404863 0.247       0.006    1.38 
## 4 Amebia~ Tulare  2001 Fema~     1     186935 0.535       0.014    2.98 
## 5 Amebia~ Los A~  2014 Fema~    17    5127242 0.332       0.193    0.531
## 6 Amebia~ Napa    2006 Total     2     131920 1.52        0.184    5.48 
## # ... with 1 more variable: Unstable <chr>

Preparing data for a merge

• We can see that the variable County in our ID data matches the variable NAME
• It’s easiest to create a new variable in our ID data to match
• We also have to subset our data to a single disease, year, and sex group

ca_id_data_formapping <- ca_id_data %>%
                         mutate(NAME = County) %>%
                         filter(Sex == "Total",Year == "2015", Disease == "Campylobacteriosis", NAME != "California")

Join data

• Now we will join our data onto the existing data in the counties2010 SPD data slot

counties2010_id<-left_join(counties2010, ca_id_data_formapping, by="NAME")
head(counties2010_id)

## Simple feature collection with 6 features and 18 fields
## geometry type:  MULTIPOLYGON
## dimension:      XY
## bbox:           xmin: -123.0143 ymin: 33.46324 xmax: -119.027 ymax: 41.99761
## epsg (SRID):    4269
## proj4string:    +proj=longlat +datum=NAD83 +no_defs
##           GEO_ID STATE COUNTY          NAME   LSAD CENSUSAREA COUNTYFP
## 1 0500000US06035    06    035        Lassen County   4541.184      035
## 2 0500000US06049    06    049         Modoc County   3917.770      049
## 3 0500000US06075    06    075 San Francisco County     46.873      075
## 4 0500000US06083    06    083 Santa Barbara County   2735.085      083
## 5 0500000US06091    06    091        Sierra County    953.214      091
## 6 0500000US06095    06    095        Solano County    821.765      095
##   STATEFP            Disease        County Year   Sex Count Population
## 1      06 Campylobacteriosis        Lassen 2015 Total     3      30969
## 2      06 Campylobacteriosis         Modoc 2015 Total     1       9507
## 3      06 Campylobacteriosis San Francisco 2015 Total   518     863108
## 4      06 Campylobacteriosis Santa Barbara 2015 Total   113     444900
## 5      06 Campylobacteriosis        Sierra 2015 Total     1       3147
## 6      06 Campylobacteriosis        Solano 2015 Total   116     429267
##     Rate CI. lower CI.upper Unstable                       geometry
## 1  9.687     1.998   28.307        * MULTIPOLYGON (((-119.9987 4...
## 2 10.519     0.266   58.592        * MULTIPOLYGON (((-120.4882 4...
## 3 60.016    54.959   65.412     <NA> MULTIPOLYGON (((-123.0139 3...
## 4 25.399    20.933   30.536     <NA> MULTIPOLYGON (((-119.7898 3...
## 5 31.776     0.805  176.918        * MULTIPOLYGON (((-120.0031 3...
## 6 27.023    22.330   32.410     <NA> MULTIPOLYGON (((-122.2667 3...

Package tmap

• tmap stands for thematic map
• It’s a powerful toolkit for creating maps with sp objects yet, with less code than the alternatives
• Syntax should be familiar to ggplot2 users, but simpler. Relatively easy to create & save interactive maps
• Install package using install.packages(“tmap”)

#install.packages("tmap")
library(tmap)

qtm mapping with tmap

• qtm() is a quick thematic map that can be made with one function call

qtm(counties2010_id)

qtm with fills

• can use fills

qtm(counties2010_id,
    fill = "Rate")

Colors!

• Use the RColorBrewer package to access some great color schemes
• install package using install.packages(“RColorBrewer”)

#install.packages("RColorBrewer")
library(RColorBrewer)
display.brewer.all()# show all ColorBrewer ramps

• And can use this ‘Shiny App’ that is part of tmap to explore the colors

#webshot::install_phantomjs()
#install.packages("shiny")
#install.packages("shinyjs")
library(shiny)
library(shinyjs)
tmaptools::palette_explorer()

qtm with colors

• can use fills and colors

qtm(counties2010_id, 
    fill = "Rate", 
    fill.palette = "Greens")

Can create a static map

• adjust tmap_mode()

tmap_mode("plot") # set tmap to static view mode
qtm(counties2010_id, 
    fill = "Rate", 
    fill.palette = "Greens") # Not Interactive

Can create an interactive map

• adjust tmap_mode()

tmap_mode("view") # set tmap to interactive view mode
qtm(counties2010_id, 
    fill = "Rate", 
    fill.palette = "Greens") # interactive

Can make a more complex map with customized popups

• We can make more complex maps using other tmap functions.
  
• We want to see polygons so here we will use tm_shape()
• Here we will also save our map to an object (map1) so we can call it later

map1 <- tm_shape(counties2010_id) + 
  tm_polygons(col = "Rate",
              id="NAME",
              palette = "Greens",
              title = "Campylobacteriosis Rate 2015",
              popup.vars = c("Disease:" = "Disease", "Year:" = "Year", "Rate Per 100,000:" = "Rate"))
map1

Basemaps

• Here are available basemaps http://leaflet-extras.github.io/leaflet-providers/preview/
• More documentation on leaflet for R: https://cran.r-project.org/web/packages/leaflet/leaflet.pdf
• And even more: https://rstudio.github.io/leaflet/

Can change basemaps

-Here we will change the baselayer

map2 <- tm_basemap(leaflet::providers$Esri.WorldStreetMap)+
  tm_shape(counties2010_id) + 
  tm_polygons(col = "Rate",
              id="NAME",
              palette = "Greens",
              title = "Campylobacteriosis Rate 2015",
              popup.vars = c("Disease:" = "Disease", "Year:" = "Year", "Rate Per 100,000:" = "Rate"))
map2

Practice Changing Colors and Basemaps

• Recreate the Campylobacteriosis Map and name the object map3
• Change the Colors and the basemap

Practice Answer

map3 <- tm_basemap(leaflet::providers$Stamen.Watercolor) +
  tm_shape(counties2010_id) + 
  tm_polygons(col = "Rate",
              id="NAME",
              palette = "Purples",
              title = "Campylobacteriosis Rate 2015",
              popup.vars = c("Disease:" = "Disease", "Year:" = "Year", "Rate Per 100,000:" = "Rate"))
map3

Can save the map as static image files/pdfs

tmap_save(map1, "California_Campylobacteriosis_Map_2015.pdf", height=10) # Static image file
tmap_save(map1, "California_Campylobacteriosis_Map_2015.png", height=6) # Static image file

Can save the map as interactive html

tmap_save(map2, "California_Campylobacteriosis_Map_2015.html") # interactive web map

Mapping with Leaflet!!

• Leaflet offers way more customization
• install package using install.packages(“leaflet”)

#install.packages("leaflet")
library(leaflet)

pal <- colorNumeric("Reds", counties2010_id$Rate)
lflt_map1 <- leaflet(counties2010_id) %>%
  addProviderTiles("Esri.WorldGrayCanvas") %>% 
  addPolygons(data=counties2010_id, 
              color = "white", 
              weight = 1, 
              opacity = 0.5,
              fillColor = ~pal(Rate),
              fillOpacity = 0.75, 
              popup = paste0("<b>County: </b>", counties2010_id$County,"<br>",
                             "<b>Disease: </b>", counties2010_id$Disease,"<br>",                             
                             "<b>Year: </b>", counties2010_id$Year,"<br>",
                             "<b>Rate: </b>", counties2010_id$Rate, " per 100,000")) %>%
  addLegend(position = "topright",
            pal = pal, 
            values = ~Rate,
            opacity = 0.75, 
            title="Rate per 100,000")

lflt_map1

Moving on to Plotting

• The plot() function in baseR can do a lot
• That said, ggplot() from the ggplot2 package (part of tidyverse) is incredibly versatile and easy to learn: https://ggplot2.tidyverse.org/

Simple bar chart

• Lets look at a bar chart of the top 10
• We can use the same dataset we created earlier

top_plot <- ggplot((ca_id_data_cleaned[1:5,cols] %>% arrange(-incidence_rate)), aes(x = Disease, y = incidence_rate)) +
  geom_col() +
  ggtitle("Top 5 Diseases in California by Incidence Rate, 2015")
top_plot

Simple line plot

• Let’s look at Campylobacteriosis over time in all of California
• First we have to prepare our data

ca_id_data_campylobacteriosis_forplot <- ca_id_data %>%
    filter(Sex == "Total", County == "California", Disease == "Campylobacteriosis") %>%
    select(County, Sex, Year, Disease, Rate)

head(ca_id_data_campylobacteriosis_forplot, 5)

## # A tibble: 5 x 5
##   County     Sex    Year Disease             Rate
##   <chr>      <chr> <dbl> <chr>              <dbl>
## 1 California Total  2009 Campylobacteriosis  15.5
## 2 California Total  2005 Campylobacteriosis  12.9
## 3 California Total  2001 Campylobacteriosis  16.0
## 4 California Total  2014 Campylobacteriosis  19.8
## 5 California Total  2012 Campylobacteriosis  21.0

Campylobacteriosis line plot

• Now we will use ggplot to create a simple line plot with a title
• Note that instead of using a pipe (%>%) ggplot uses a plus (+) to string lines of code/arguments together

campy_plot <- ggplot(ca_id_data_campylobacteriosis_forplot, aes(x = Year,  y = Rate)) +
  geom_line(aes(), size = 1) +
  ggtitle("California Campylobacteriosis Incidence Rate, 2001-2015")
campy_plot

Compare with other Diseases

• Let’s compare rates of multiple foodborne diseases over time in all of California
• First we have to prepare our data

ca_id_data_foodborne_forplot <- ca_id_data %>%
  filter(Sex == "Total", County == "California", Disease %in% c("Campylobacteriosis", "Salmonellosis", "Shigellosis", "Giardiasis", "Cryptosporidiosis", "E. coli O157"))%>%
  select(County, Sex, Year, Disease, Rate)

head(ca_id_data_foodborne_forplot, 5)

## # A tibble: 5 x 5
##   County     Sex    Year Disease             Rate
##   <chr>      <chr> <dbl> <chr>              <dbl>
## 1 California Total  2009 Campylobacteriosis  15.5
## 2 California Total  2005 Campylobacteriosis  12.9
## 3 California Total  2001 Campylobacteriosis  16.0
## 4 California Total  2014 Campylobacteriosis  19.8
## 5 California Total  2012 Campylobacteriosis  21.0

Foodborne Disease line plot

• In the last plot, the X axis only had labels for 2004, 2008 and 2012 so we will add in breaks for every year.
• We will also adjust the angle of the X axis labels
• We will also add in axis labels

foodborne_plot <- ggplot(ca_id_data_foodborne_forplot, aes(x = Year, y = Rate)) +
  geom_line(aes(color = Disease), size = 1) +
  scale_x_continuous(breaks = seq(2001, 2015, 1)) +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)) +
  labs(x = "Year", y = "Incidence Rate (Per 100,000)") +
  ggtitle("California Foodborne Disease Incidence Rates, 2001-2015")
foodborne_plot

Interactive Foodborne Disease line plot!!

• Can use the plotly package to make the plot interactive
• Install the package using install.packages(“plotly”)

#install.packages("plotly")
library(plotly)
ggplotly(foodborne_plot)

Can save the plot as static image files/pdfs

• Use the ggsave() function

ggsave(file = "Cal_Foodborne_2001-2015.pdf",
       plot = foodborne_plot, # or give ggplot object name as in myPlot,
       width = 5, 
       height = 5, 
       units = "in", # other options c("in", "cm", "mm"), 
       dpi = 300)

ggsave(file = "Cal_Foodborne_2001-2015.png",
       plot = foodborne_plot, # or give ggplot object name as in myPlot,
       width = 5, 
       height = 5, 
       units = "in", # other options c("in", "cm", "mm"), 
       dpi = 300)

We can also save the interactive plotly-wrapped plot in html

• Need to use the htmlwidgets package. Install the package using install.package(“htmlwidgets”)

#install.package("htmlwidgets")
library(htmlwidgets)
plotly_foodborne_plot <- ggplotly(foodborne_plot)
htmlwidgets::saveWidget(as_widget(plotly_foodborne_plot), "Cal_Foodborne_2001-2015.html")

Dig in to the Viz!

• We see that Campylobacteriosis has increased dramatically. Which counties are driving that?
• To answer that question, we might want to look at Campylobacteriosis by County for 2015
• In the past we have prepared a dataset to feed into ggplot, but we can use () to do the wrangling from within the function!

campy_2015_plot <- ggplot(data = (ca_id_data %>%
                                    filter(Sex == "Total", Year == "2015", Disease == "Campylobacteriosis", County != "California")),
        aes(x = reorder(County,-Rate), y = Rate))+
        geom_col(aes())+
        theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5))+
        labs(x = "County", y = "Incidence Rate (Per 100,000)")+
        ggtitle("California Campylobacteriosis Rate by County, 2015")
campy_2015_plot

ggplotly(campy_2015_plot)

There are many great resources for plotting in R.

• This one is especailly valuable becuase of all the great theory and practice for making charts, and examples of R code for many many type of visualization https://www.data-to-viz.com/

Advanced Topic–Accessing Census API with TidyCensus

Shift to accessing Census Data

• There are a few great packages for accessing Census API data.
  
• The easiest one to use is the tidycensus package
• You can install using install.packages(“tidycensus”)

#install.packages("tidycensus")
library(tidycensus)

# Load census key - Get a key at https://api.census.gov/data/key_signup.html and change it below!
my_census_api_key <- "39503dad5b1feef439deebd1a721c7a4aa4bee04"

# Make tidycensus aware of census key.  Also, To install your API key for use in future sessions, run this function with `install = TRUE`.
census_api_key(my_census_api_key) 

Set up the API pull

• Now we have loaded our key we need to figure out which data we want to call
• We need to specify geography and the variables we want
• In this case, we want to get data at the County level, but it’s also available at other census geographies (i.e. tract, block groups, etc.): https://walkerke.github.io/tidycensus/articles/basic-usage.html

#Identify state(s) and county(s) or place(s) of interest.  YOu can find all available geographies and calls here:https://walkerke.github.io/tidycensus/articles/basic-usage.html
my_states <- c("06") # CA

#load the variables table to find what you want to look at/fetch
cenvar_table <-load_variables(year=2017, dataset = "acs5", cache=T)
B17001 <- filter(cenvar_table, str_detect(name, "B17001"))

#Take a look at the data frame to find the names of census tables. This will give you the table and variable codes. Use "View(cenvar_table)" and filter within the table to get the codes for a specific variable
View(B17001) 


#now you can identify a single variable or create a vector of variables
pop_total <- "B17001_001E" #can specify a single variable to pull
pop_poverty <- "B17001_002E" #can specify a single variable to pull

Pull Data from Census API using get_acs()

#Fetch Data of Interest.  Note some of the specifications here.  You specify the geometry.
## Please note: `get_acs()` now defaults to a year or endyear of 2017.

#here I pull tract level data for only the two vars we specified above (B17001_001E and B17001_002E) for Alameda County, CA 
county_poverty_acs5_2017 <- get_acs(geography = "county", 
                              variables = c(pop_total,pop_poverty), 
                              year=2017, survey="acs5",
                              state = my_states,
                              geometry = F)


#explore dataframe, it's long
head(county_poverty_acs5_2017, 3)

## # A tibble: 3 x 5
##   GEOID NAME                       variable   estimate   moe
##   <chr> <chr>                      <chr>         <dbl> <dbl>
## 1 06001 Alameda County, California B17001_001  1602357  1188
## 2 06001 Alameda County, California B17001_002   181194  4374
## 3 06003 Alpine County, California  B17001_001     1188   179

Process our ACS data to join it with our map!

• Now we want a map of the % of the population that is in poverty by county
• We need to prepare our data and create a new variable with the percentage that we want to map

# Select the columns of interest 
county_poverty_acs5_2017_wide <- county_poverty_acs5_2017 %>%
                                select("GEOID","variable","estimate") %>%
                                spread(key = variable, value = estimate)


# Rename columns
colnames(county_poverty_acs5_2017_wide) <- c("county", "pop_total", "pop_poverty")

county_poverty_acs5_2017_wide <- county_poverty_acs5_2017_wide %>%
                            mutate(pct_poverty = round((pop_poverty / pop_total) * 100, 1)) #create a variable with the % of the population that is under 18.

Create a Join Variable in our Map Data

• Since we will want to join this with our county map data, we need to identify which variable to use to join. Unfortunately there isn’t a native variable that is common to both datasets
• We need to use the substr() function to extract out the final 5 digits from the GEO_ID variable in our County level simple features dataframe

head(counties2010)

## Simple feature collection with 6 features and 8 fields
## geometry type:  MULTIPOLYGON
## dimension:      XY
## bbox:           xmin: -123.0143 ymin: 33.46324 xmax: -119.027 ymax: 41.99761
## epsg (SRID):    4269
## proj4string:    +proj=longlat +datum=NAD83 +no_defs
##             GEO_ID STATE COUNTY          NAME   LSAD CENSUSAREA
## 187 0500000US06035    06    035        Lassen County   4541.184
## 188 0500000US06049    06    049         Modoc County   3917.770
## 189 0500000US06075    06    075 San Francisco County     46.873
## 190 0500000US06083    06    083 Santa Barbara County   2735.085
## 191 0500000US06091    06    091        Sierra County    953.214
## 192 0500000US06095    06    095        Solano County    821.765
##                           geometry COUNTYFP STATEFP
## 187 MULTIPOLYGON (((-119.9987 4...      035      06
## 188 MULTIPOLYGON (((-120.4882 4...      049      06
## 189 MULTIPOLYGON (((-123.0139 3...      075      06
## 190 MULTIPOLYGON (((-119.7898 3...      083      06
## 191 MULTIPOLYGON (((-120.0031 3...      091      06
## 192 MULTIPOLYGON (((-122.2667 3...      095      06

counties2010_acs <- counties2010 %>% 
  mutate(county = substr(GEO_ID,10,14))

head(counties2010_acs)

## Simple feature collection with 6 features and 9 fields
## geometry type:  MULTIPOLYGON
## dimension:      XY
## bbox:           xmin: -123.0143 ymin: 33.46324 xmax: -119.027 ymax: 41.99761
## epsg (SRID):    4269
## proj4string:    +proj=longlat +datum=NAD83 +no_defs
##           GEO_ID STATE COUNTY          NAME   LSAD CENSUSAREA
## 1 0500000US06035    06    035        Lassen County   4541.184
## 2 0500000US06049    06    049         Modoc County   3917.770
## 3 0500000US06075    06    075 San Francisco County     46.873
## 4 0500000US06083    06    083 Santa Barbara County   2735.085
## 5 0500000US06091    06    091        Sierra County    953.214
## 6 0500000US06095    06    095        Solano County    821.765
##                         geometry COUNTYFP STATEFP county
## 1 MULTIPOLYGON (((-119.9987 4...      035      06  06035
## 2 MULTIPOLYGON (((-120.4882 4...      049      06  06049
## 3 MULTIPOLYGON (((-123.0139 3...      075      06  06075
## 4 MULTIPOLYGON (((-119.7898 3...      083      06  06083
## 5 MULTIPOLYGON (((-120.0031 3...      091      06  06091
## 6 MULTIPOLYGON (((-122.2667 3...      095      06  06095

Join ACS data with the SF mapping data

counties2010_acs_joined<-left_join(counties2010_acs, county_poverty_acs5_2017_wide, by="county")

Map it!

map_acs<-tm_shape(counties2010_acs_joined) + 
  tm_polygons(col = "pct_poverty",
              id="NAME",
              palette = "Reds",
              title = "2013-2017 Population Poverty Rate (%)",
              popup.vars = c("Percent of Population in Poverty" = "pct_poverty", "Total Population:" = "pop_total", "Population in Poverty" = "pop_poverty"))

map_acs

Advanced Topic–Shiny Apps

• We’ve also developed a simple Shiny App for you to see
• First we need to send some of our data objects to the app

saveRDS(ca_id_data, file = "CCLHDME_R_shiny_app/data/ca_id_data.rds")

• Now we need to open the app to run it
• 1. Navigate to File and Click
• 2. Under File, Select Open File
• 3. When the file browser opens, Open the folder **CCLHDME_R_shiny_app*
• 4. Select app.R
• 4. Click Open
• Note on publishing Shinyapps:
• In order to publish a shinyapp, you need to have: R Studio Connect, R Shiny Server (there is a free version or a Pro/Enterprise version) or cloud-hosting service such as shinyapps.io (there’s a free version and paid versions)

Other help resources

• Great cheatsheets, tutorials, documentation/manuals, entire books and Error Support online. Here is a sampling of things we’ve used:
• Cheatsheets:
• https://www.rstudio.com/resources/cheatsheets/
• Tutorials:
• https://www.rstudio.com/online-learning/
• https://stats.idre.ucla.edu/
• https://www.coursera.org/learn/r-programming
• https://www.lynda.com/R-training-tutorials/1570-0.html
• https://www.datacamp.com/
• https://www.datacamp.com/courses/free-introduction-to-r
• https://campus.datacamp.com/courses/introduction-to-the-tidyverse/data-wrangling-1?ex=1
• https://www.datacamp.com/courses/data-visualization-with-ggplot2-1
• Documentation/Manuals:
• https://cran.r-project.org/doc/manuals/R-intro.pdf
• https://style.tidyverse.org/
• Books:
• https://r4ds.had.co.nz/
• https://tbrieder.org/epidata/course_reading/e_aragon.pdf
• Error Support:
• https://stackoverflow.com/questions/tagged/r

Notes on Workshop Development

• Special thanks to UC Berkeley D-Lab for providing inspiration for this workhop and instructive style: https://dlab.berkeley.edu/
• Also special thanks to the R Community for developing instructive tools that we drew upon heavily while designing this workshop

That’s it for now…

• Thank you for your participation
• We welcome your feedback about the workshop!
• Let us know if you have any questions!